Anomaly Detection with Pattern of Life Extraction for GMTI Tracking

Ground Moving Target Indicator (GMTI) uses the concept of airborne surveillance of moving ground objects to observe and take actions if necessary. With the improvement of surveillance technology, tracking individual target from the surveillance became possible, which allows the extraction of useful features for advance usage. Such features, known as tracks, are the results of GMTI tracking. The system for this paper will be based on GMTI track data. Though the quality of the tracker plays a crucial role to the system performance of this paper, the development of the tracker will not be discuss in this paper. The system will use simulated ideal GMTI tracks as inputs. This paper presents Pattern of Life (PoL) extraction and Anomaly Detection System (ADS). The results from PoL extraction will be used to improve the performance of ADS. The proposing ADS is a semi-supervised learning detection system, in which the system takes prior information to support and improve detection performance, but will still operate without prior information. The results from ADS will also be evaluated. The ADS will use a combination of various anomaly detection algorithms for different anomaly events including statistical approach using Gaussian Mixture Model Expectation Maximization (GMM-EM), Hidden Markov Model (HMM), graphical approach using Weiler-Atherton Polygon Clipping (WAPC) and various clustering algorithms such as K-mean clustering, Spectral clustering and DBSCAN.